The relationship between risk perception, anxiety and paranoia - A predictive model in a community sample.

Background: Biases in risk perception (e.g. excessive attribution of likelihood of negative events happening to oneself, or perceived harm of neutral events) have been suggested as risk factors for psychopathologies such as generalised anxiety and persecutory ideation, although this line of research is limited by small samples and a lack of a suitable risk perception scale. Methods: Using the Risk Perception Questionnaire, four risk perception dimensions (likelihood, harm, controllability, and intentionality) of negative and neutral events were tested in association with anxiety and paranoia. In view of common co-occurrence between the two symptom variables, their associations with risk perception were tested by using partial correlations (at baseline) and comparisons of cross-lagged path models (over 3 months). Results: A representative community-based sample of 445 adults were included. At baseline, after controlling for correlations between levels of anxiety and paranoia, anxiety was uniquely correlated with three risk perception dimensions for negative events (likelihood, harm, and intentionality), whereas paranoia was uniquely correlated with all risk perception dimensions for both negative and neutral events. The best-fitted cross-lagged path model revealed that, after controlling for auto-regressions within variables, baseline level of anxiety predicted perceived harm of negative events at 3 months, whereas baseline levels of perceived intentionality of neutral events and likelihood of negative events predicted level of paranoia at 3 months. Conclusions: While risk perception of negative events is shared between anxiety and paranoia, risk perception of neutral events is uniquely characteristic of paranoia. Implications on maintenance of sub-clinical symptoms are discussed.